Abstract: To calculate the frequency distribution parameters of annual precipitation with different frequency distributions, traditional estimation methods need to derive complicated formulas. This paper uses new intelligent optimization algorithms for estimating such parameters to improve estimation accuracy. Three algorithms, self-adaptive differential evolution algorithm based on opposition-based learning (OL-ADE), dragonfly algorithm (DA), and hybrid genetic and particle swarm algorithm (HGAPSO), are applied to calculation of the distribution parameters according to the optimization criteria for the annual precipitation series of the meteorological stations of Meixian, Fengxian, and Fengxiang in Shaanxi. And estimation results of the three above mentioned methods and three traditional methods are evaluated using the technique for order preference by similarity to ideal solution (TOPSIS). The results show that compared to traditional methods, three intelligent optimization algorithms fit better with the annual precipitation frequency distribution parameters, the accuracy of HGAPSO is best, and DA and OL-ADE are comparable.

Key words: parameter estimation, self-adaptive differential evolution algorithm based on opposition-based learning, dragonfly algorithm, hybrid genetic and particle swarm algorithm, TOPSIS evaluation method